检索结果-内蒙古大学图书馆

visual classification and Detection of Power Inspection Images Based on Federated Learning

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS 2024年第4期60卷 5460-5469页

作者： Zhong, Linlin Liu, Keyu Southeast Univ Sch Elect Engn Nanjing 210096 Peoples R China Southeast Univ SEU Monash Joint Grad Sch Suzhou 215123 Peoples R China

The power lines and equipment of power system are inspected regularly by Unmanned Aerial Vehicles (UAVs) and video monitoring devices, which generates large quantities of power inspection images. The deep learning (DL) methods, such as visual classification and detection models, can process such images effectively. However, due to the data privacy regulations, the inspection images collected by a power company are not allowed to be shared with others. The data from a single owner is limited not only in quantity but also in type, which cannot always support to train a high-performance model. In this work, we propose a federated learning (FL) based method for processing power inspection images, which allows different data owners to cooperatively train visual classification and detection models without sharing their local images. To improve the training efficiency, we further propose a Federated Round-level Momentum (FedRM) method by adding a momentum term during the aggregation of model weights. We demonstrate the proposed method in three real-world power inspection image datasets for visual classification, object detection and defect detection tasks respectively, and the effects of non-independent and identically distributed (Non-IID) data on the FL based models are investigated. The results show that the performance of the FL models for power inspection images is higher than that of local trained models. It is also verified that our proposed FedRM method improves the FL-training efficiency significantly, which could reach at most 5x, 3.7x, and 6.8x speedup on visual classification, object detection, and defect detection tasks respectively.

关键词： Federated learning non-independent and identically distributed (Non-IID) object detection power inspection visual classification Federated learning non-independent and identically distributed (Non-IID) object detection power inspection visual classification

来源：评论

学校读者我要写书评

暂无评论

Multi-granularity hypergraph-guided transformer learning framework for visual classification

引用

visual COMPUTER 2025年第4期41卷 2391-2408页

作者： Jiang, Jianjian Chen, Ziwei Lei, Fangyuan Xu, Long Huang, Jiahao Yuan, Xiaochen Guangdong Polytech Normal Univ Guangdong Prov Key Lab Intellectual Property & Big Guangzhou 510665 Peoples R China Macao Polytech Univ Fac Appl Sci Macau 999078 Peoples R China

Fine-grained single-label classification tasks aim to distinguish highly similar categories but often overlook inter-category relationships. Hierarchical multi-granularity visual classification strives to categorize image labels at various hierarchy levels, offering optimize label selection for people. This paper addresses the hierarchical multi-granularity classification problem from two perspectives: (1) effective utilization of labels at different levels and (2) efficient learning to distinguish multi-granularity visual features. To tackle these issues, we propose a novel multi-granularity hypergraph-guided transformer learning framework (MHTL), seamlessly integrating swin transformers and hypergraph neural networks for handling visual classification tasks. Firstly, we employ swin transformer as an image hierarchical feature learning (IHFL) module to capture hierarchical features. Secondly, a feature reassemble (FR) module is applied to rearrange features at different hierarchy levels, creating a spectrum of features from coarse to fine-grained. Thirdly, we propose a feature relationship mining (FRM) module, to unveil the correlation between features at different granularity. Within this module, we introduce a learnable hypergraph modeling method to construct coarse to fine-grained hypergraph structures. Simultaneously, multi-granularity hypergraph neural networks are employed to explore grouping relationships across different granularities, thereby enhancing the learning of semantic feature representations. Finally, we adopt a multi-granularity classifier (MC) to predict hierarchical label probabilities. Experimental results demonstrate that MHTL outperforms other state-of-the-art classification methods across three multi-granularity datasets. The source code and models are released at https://***/JJJTF/MHTL.

关键词： Hierarchical multi-granularity visual classification Hypergraph neural network Swin transformer

来源：评论

学校读者我要写书评

暂无评论

Research on fine-grained visual classification based on salient object detection

引用

LASER PHYSICS 2025年第5期35卷 055201-055201页

作者： Tao, Rong Shen, Dafu Zhang, Leihong Shen, Zimin Qian, Zhenhua Xu, Banglian Zhang, Dawei Univ Shanghai Sci & Technol Sch Opt Elect & Comp Engn Shanghai 200093 Peoples R China Shanghai Jiao Tong Univ Renji Hosp Sch Med Dept Cardiovasc Surg 160 Pujian Rd Shanghai 200127 Peoples R China Soochow Univ Affiliated Hosp 1 Dept Cardiovasc Surg Suzhou 215123 Jiangsu Peoples R China Zhejiang Univ State Key Lab Extreme Photon & Instrumentat Hangzhou 310009 Peoples R China Huzhou Vocat & Tech Coll Sch Mechatron & Automot Engn Huzhou 313099 Peoples R China

In response to the current algorithm of fine-grained image classification is greatly disturbed by complex backgrounds with similar features to the objects to be classified, salient object detection is introduced into fine-grained visual classification (FGVC), and a FGVC algorithm based on saliency target detection is proposed in this paper. The target object and background of the original image are segmented using the saliency map obtained from the U(2)Net salient object detection model. The images with background interference removed are fed into the EfficientNet image classification network to achieve classification. Experiments show that this method has better classification results compared with the original image directly into EfficientNet and the EfficientNet with attention mechanism. The parameters are tuned on the base EfficientNet to obtain higher classification accuracy.

关键词： salient object detection visual classification U-2 Net

来源：评论

学校读者我要写书评

暂无评论

Edge Devices Clustering for Federated visual classification: A Feature Norm Based Framework

引用

IEEE TRANSACTIONS ON IMAGE PROCESSING 2023年 32卷 995-1010页

作者： Wei, Xiao-Xiang Huang, Hua Beijing Inst Technol Sch Comp Sci & Technol Beijing 100081 Peoples R China Beijing Normal Univ Sch Artificial Intelligence Beijing 100875 Peoples R China

Federated learning is a privacy-preserving distributed learning paradigm where multiple devices collaboratively train a model, which is applicable to edge computing environments. However, the non-IID data distributed in multiple devices degrades the performance of the federated model due to severe weight divergence. This paper presents a clustered federated learning framework named cFedFN for visual classification tasks in order to reduce the degradation. Especially, this framework introduces the computation of feature norm vectors in the local training process and divides the devices into multiple groups by the similarities of the data distributions to reduce the weight divergences for better performance. As a result, this framework gains better performance on non-IID data without leakage of the private raw data. Experiments on various visual classification datasets demonstrate the superiority of this framework over the state-of-the-art clustered federated learning frameworks.

关键词： Data models Feature extraction Computational modeling visualization Training Federated learning Adaptation models edge computing visual classification feature norm clients clustering

来源：评论

学校读者我要写书评

暂无评论

Reconciliation of statistical and spatial sparsity for robust visual classification

引用

NEUROCOMPUTING 2023年第1期529卷 140-151页

作者： Cheng, Hao Yap, Kim-Hui Wen, Bihan Nanyang Technol Univ Sch Elect & Elect Engn Singapore 639798 Singapore

Recent image classification algorithms, by learning deep features from large-scale datasets, have achieved significantly better results comparing to the classic feature-based approaches. However, there are still various challenges of image classifications in practice, such as classifying noisy image or image-set queries, and training deep image classification models over the limited-scale dataset. Instead of applying generic deep features, the model-based approaches can be more effective and data-efficient for robust image and image-set classification tasks, as various image priors are exploited for modeling the inter -and intra-set data variations while preventing over-fitting. In this work, we propose a novel Joint Statistical and Spatial Sparse representation scheme, dubbed J3S, to model the image or image-set data for classification. J3S utilized joint sparse representation to reconcile both the local image structures and global Gaussian distribution mapped into Riemannian manifold. The learned J3S models are used for robust image and image-set classification tasks. Experiments show that the proposed J3S-based image classification scheme outperforms the popular or state-of-the-art competing methods over FMD, UIUC, ETH-80 and YTC databases.(c) 2023 Elsevier B.V. All rights reserved.

关键词： Sparse representation Gaussian model Riemannian manifold Dictionary learning visual classification

来源：评论

学校读者我要写书评

暂无评论

U-SPDNet: An SPD manifold learning-based neural network for visual classification

引用

NEURAL NETWORKS 2023年第1期161卷 382-396页

作者： Wang, Rui Wu, Xiao-Jun Xu, Tianyang Hu, Cong Kittler, Josef Jiangnan Univ Sch Artificial Intelligence & Comp Sci Wuxi 214122 Peoples R China Jiangnan Univ Jiangsu Prov Engn Lab Pattern Recognit & Computat Wuxi 214122 Peoples R China Univ Surrey Ctr Vis Speech & Signal Proc CVSSP Guildford GU2 7XH England

With the development of neural networking techniques, several architectures for symmetric positive definite (SPD) matrix learning have recently been put forward in the computer vision and pattern recognition (CV&PR) community for mining fine-grained geometric features. However, the degradation of structural information during multi-stage feature transformation limits their capacity. To cope with this issue, this paper develops a U-shaped neural network on the SPD manifolds (U-SPDNet) for visual classification. The designed U-SPDNet contains two subsystems, one of which is a shrinking path (encoder) making up of a prevailing SPD manifold neural network (SPDNet (Huang and Van Gool, 2017)) for capturing compact representations from the input data. Another is a constructed symmetric expanding path (decoder) to upsample the encoded features, trained by a reconstruction error term. With this design, the degradation problem will be gradually alleviated during training. To enhance the representational capacity of U-SPDNet, we also append skip connections from encoder to decoder, realized by manifold-valued geometric operations, namely Riemannian barycenter and Riemannian optimization. On the MDSD, Virus, FPHA, and UAV-Human datasets, the accuracy achieved by our method is respectively 6.92%, 8.67%, 1.57%, and 1.08% higher than SPDNet, certifying its effectiveness.(c) 2022 Published by Elsevier Ltd.

关键词： SPD manifold Neural network visual classification Skip connection Riemannian barycenter Riemannian optimization

来源：评论

学校读者我要写书评

暂无评论

Learning Representation on Optimized High-Order Manifold for visual classification

引用

IEEE TRANSACTIONS ON MULTIMEDIA 2022年 24卷 3989-4001页

作者： Ma, Xueqi Liu, Weifeng Tian, Qi Gao, Yue Tsinghua Univ Sch Software KLISS BNRistTHUIBCS Beijing 100084 Peoples R China Xidian Univ State Key Lab Integrated Serv Networks Xian 710126 Peoples R China China Univ Petr East China Coll Control Sci & Engn Qingdao 266580 Peoples R China Huawei Cloud & AI Shenzhen 518129 Peoples R China

Graph convolutional networks (GCNs) and graph neural networks (GNNs) have demonstrated convincing performance on many tasks by learning the intrinsic structure of the data. However, it is still valuable and challenging to consider the complex and complete correlations of objects, i.e., high-order manifold structures, for representation learning. In this paper, we present a novel representation learning method that utilizes the optimized high-order manifold of the data for classification tasks of nonstructural data and graph-structure data. In the method, we fully explore the complicated relationship of samples by highlighting the high-order manifold information in a hypergraph. Specifically, we incorporate high-order manifold information by graph p-Laplacian into a hypergraph and propose p-Laplacian-based hypergraph neural networks (pLapHGNN) to significantly learn hidden layer representations that encode both the high-order structure of data and the high-order manifold geometrical information. Confronting the difficulties of obtaining optimized high-order manifolds of the data, we propose an effective approximate approach by graph p-Laplacian representing the relationship of hyperedges in the hypergraph. Furthermore, we study the weights of hyperedges in a hypergraph with high-order manifold information. Experiments on the ModelNet40 dataset and NTU dataset demonstrate that the proposed method is more effective than the other popular methods for 3D shape recognition. Extensive experiments on other visual classification tasks and citation networks also show the superiority of our proposed method for representation learning.

关键词： High-order manifold hypergraph graph neural networks visual classification

来源：评论

学校读者我要写书评

暂无评论

Discriminative elastic-net broad learning systems for visual classification

引用

APPLIED SOFT COMPUTING 2024年 155卷

作者： Li, Yanting Jin, Junwei Geng, Yun Xiao, Yang Liang, Jing Chen, C. L. Philip Zhengzhou Univ Light Ind Sch Comp & Commun Engn Zhengzhou 450001 Peoples R China Henan Univ Technol Key Lab Grain Informat Proc & Control Minist Educ Zhengzhou Peoples R China Henan Prov Key Lab Grain Photoelect Detect & Contr Zhengzhou Peoples R China Henan Univ Technol Sch Artificial Intelligence & Big Data Zhengzhou 450001 Peoples R China Zhengzhou Univ Sch Elect & Informat Engn Zhengzhou 450001 Peoples R China Univ Alabama Dept Comp Sci Tuscaloosa AL 35487 USA South China Univ Technol Sch Comp Sci & Engn Guangzhou Peoples R China

The broad learning system (BLS) has garnered significant attention in the realm of visual classification due to its exceptional balance between accuracy and efficiency. However, the supervision mechanism in BLS typically relies on strict binary labels, limiting the approximation freedom and failing to represent the data distribution adequately. Furthermore, the inadequacy of the guidance mechanism for the output weights hinders the precise approximation of the target spaces by the input features. To tackle these issues, in this paper, we propose two discriminative elastic -net regularized BLS models with label enhancement, achieving the following significant objectives: Firstly, the proposed label enhancement technologies can substantially augment the margins between different classes while enhancing the diversity within label spaces. Secondly, the compactness and effectiveness of the output weights can be further improved via the guidance of elasticnet regularization of singular values. Thirdly, our proposed algorithms can be efficiently optimized with the augmented Lagrangian method, whose convergence and calculation complexity can be guaranteed well with solid theoretical analysis. Extensive experiments are intended on various popular databases to compare our proposed models with numerous other state-of-the-art recognition algorithms. The numerical results indicate that our proposed algorithms can achieve the best face recognition accuracy of 99.67%, and 97.54% for object recognition. Even in challenging recognition tasks, our methods can still yield an average improvement of 0.8 percentage points.

关键词： Broad learning system Label enhancement Elastic-net regularization Discriminative visual classification

来源：评论

学校读者我要写书评

暂无评论

Fisher regularized discriminative broad learning system for visual classification

引用

APPLIED SOFT COMPUTING 2024年 167卷

作者： Li, Xianghua Wei, Jinlong Jin, Junwei Xu, Tao Yu, Dengxiu Northwestern Polytech Univ Sch Artificial Intelligence OPt & Elect iOPEN Xian 710072 Peoples R China Henan Univ Technol Sch Artificial Intelligence & Big Data Zhengzhou 450001 Peoples R China Northwestern Polytech Univ Unmanned Syst Res Inst Xian 710072 Peoples R China

The Broad Learning System (BLS) is an innovative learning paradigm with significant success in image classification. However, the 0-1 labeling matrix employed in BLS struggles to align with the true data distribution, limiting the flexibility of the regression objective. The l(2)-norm-based Broad Learning System (L2DBLS) introduces the epsilon-dragging technique to enhance labeling diversity, but the randomness inherent in epsilon-dragging weakens the label correlation within the same category. This paper proposes the Fisher Regularized Discriminative Broad Learning System (FRBLS) to tackle these issues and aims to achieve the following objectives: Firstly, the generated label matrix offers sufficient flexibility to maintain intra-class compactness and inter-class separability. Secondly, the constraints of Fisher Regularization on the same class of features ensure better alignment between samples and labels. Finally, the overall solution process is optimized using an ADAM- based alternating multiplier method, which ensures closed-form solutions at each iteration. Experimental results demonstrate that FRBLS achieves up to 98% accuracy on various face and object datasets, offering superior time efficiency and a 1% improvement in classification performance compared to recent state-of-the-art methods.

关键词： Broad learning system Fisher regularized Label dragging visual classification Regression labels

来源：评论

学校读者我要写书评

暂无评论

Autolabeling-Enhanced Active Learning for Cost-Efficient Surface Defect visual classification

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2021年 70卷 1-15页

作者： Yang, Hua Song, Kaiyou Mao, Fangqin Yin, Zhouping Huazhong Univ Sci & Technol Sch Mech Sci & Engn State Key Lab Digital Mfg Equipment & Technol Wuhan 430074 Hubei Peoples R China

Active learning can reduce the human effort required for labeling training samples while preserving the performance of visual classifiers. However, existing active learning frameworks cannot be used to perform visual classification of industrial product surface defects because they still require intensive manual annotation efforts. In this study, we propose a cost-efficient autolabeling-enhanced active learning (ALEAL) framework to reduce the human annotation effort required for surface defect visual classification. The proposed ALEAL framework employs a deep convolutional neural network (CNN) as a visual classifier trained from an initial set of human-labeled training samples. Then, the collected unlabeled training samples are input into the classifier for category confidence estimation. Next, a novel diverse cost-effective query strategy (DCEQS) is proposed to select some high-confidence samples for autolabeling and some informative samples for sample proposals that need labeling. Subsequently, to further reduce the human annotation effort, a novel autolabeling module is proposed and introduced in ALEAL that can automatically label a portion of the informative unlabeled training samples selected by the DCEQS. In this study, a novel attention-based similarity measurement network (ASMN) is proposed as an implementation of this autolabeling module by measuring the similarity between unlabeled and labeled samples. Finally, the remaining unlabeled samples are annotated by human experts, and all the newly labeled samples are used to retrain the classifier. Through the autolabeling process from the DCEQS and ASMN, ALEAL can automatically label additional training samples and achieve a competitive performance while requiring few human-labeled training samples, which is highly important in industrial applications. Extensive experimental results show that, compared with popular active learning methods, ALEAL can dramatically reduce the effort involved in human annotation an

关键词： Active learning convolutional neural network (CNN) deep learning surface defects visual classification

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：